Official C12 Clone - Build and Support Thread
- Thread starter Matador
- Start date
Help Support GroupDIY Audio Forum:
kidvybes said:
Humm, that RK and TK sure do look a lot alike. Thanks for pointing that out.
What did you perceive as the preferred difference using Chung's capsule?
I chose the RK12 for my first C12 build, in which I thought I'd try a larger output cap [maybe even a PIO], along with Chung's dual mesh custom C12 headbasket [to move a bit in the direction of an M251]. I also have a few old RCA 6072's to try out.
Matador
Well-known member
trans4funks1 said:
Most of the relevant background is here.
In short, you want to take the 12AT7 plate curves, and pick an operating point where the input and output swings are roughly the same.
This site is very useful for this. You can change the tube type to ECC81 (e.g. 12AT7), then slide the supply voltage down to 120V, and set the input swing to 500mV peak, and set the bias at -1V. If you dial the load-line up and down from where it intersects the Y axis, you'll find a point at about 2.55mA where distortion is minimized, and overall gain is roughly the same at 12AY7. 120V/2.55mA = 47k.
Interestingly enough, absolutely lowest distortion for a 12AT7 occurs at -0.66V bias (0.9% distortion), so that is worthy of experimentation as well. Idling point is 1.28mA, grid bias is -0.66V, so the cathode resistance is 536ohms, well within adjustment range of the bias pot (which can do 0 to 5K).
If you leave in the 100K, distortion is much higher at 3%, but it will work.
trans4funks1
Well-known member
- Joined
- Feb 4, 2013
- Messages
- 328
Thank you Matador.
It took me a while to figure out where to adjust some of the inputs but I was able to follow your explanation, and consider the idea that the plate resistor can be sized after an analysis of the distortion.
It took me a while to figure out where to adjust some of the inputs but I was able to follow your explanation, and consider the idea that the plate resistor can be sized after an analysis of the distortion.
chunger
Well-known member
UPDATE:
New V1.4A microphone board has been confirmed working on the first prototype build. The layout of the board is familiar and has not significantly changed, but the main improvement is we are now working with a glass reinforced hydrocarbon ceramic board material with electrical performance properties close to PTFE. Matador is the one who came up with the plan and should be able to shed more light on the technical details, but to my understanding.
- HiZ performance when coupled with our existing proprietary solder mask specification should be improved
- very low moisture absorption decreases effects of humidity on the circuit
- white board material base color will render yellow solder mask more vibrantly for superior tone
The new boards are thinner and more flexible than FR4, and there were some layout specifics that needed to be addressed by Matador. These factors required a bit more time and effort to execute than a standard revision on regular PCB's.
The test microphone was built to use the NOS Mullard 12AT7 tube. I chose to build a hybrid microphone to test in studio with an Eric Heiserman HK47 capsule. Output capacitor is 2uf. Transformer is Cinemag CM-13114.
The current kit standard 68.1K R1, R2 resistors (centered on 6072A tube) were used in the power supply. At full adjustment on the 12AT7 tube, we have 118.5V at the B+, and for now, I am ok to leave it like that because I am extremely short on PSU boards at this time as the production batch of boards has not arrived yet. 33K R1, R2 resistors will better center the trim pot to the 12AT7 tube, but I need to share this power supply with standard 6072A mics.
The mic will be passed off to our affiliate studio for testing to further verify function.
New V1.4A microphone board has been confirmed working on the first prototype build. The layout of the board is familiar and has not significantly changed, but the main improvement is we are now working with a glass reinforced hydrocarbon ceramic board material with electrical performance properties close to PTFE. Matador is the one who came up with the plan and should be able to shed more light on the technical details, but to my understanding.
- HiZ performance when coupled with our existing proprietary solder mask specification should be improved
- very low moisture absorption decreases effects of humidity on the circuit
- white board material base color will render yellow solder mask more vibrantly for superior tone
The new boards are thinner and more flexible than FR4, and there were some layout specifics that needed to be addressed by Matador. These factors required a bit more time and effort to execute than a standard revision on regular PCB's.
The test microphone was built to use the NOS Mullard 12AT7 tube. I chose to build a hybrid microphone to test in studio with an Eric Heiserman HK47 capsule. Output capacitor is 2uf. Transformer is Cinemag CM-13114.
The current kit standard 68.1K R1, R2 resistors (centered on 6072A tube) were used in the power supply. At full adjustment on the 12AT7 tube, we have 118.5V at the B+, and for now, I am ok to leave it like that because I am extremely short on PSU boards at this time as the production batch of boards has not arrived yet. 33K R1, R2 resistors will better center the trim pot to the 12AT7 tube, but I need to share this power supply with standard 6072A mics.
The mic will be passed off to our affiliate studio for testing to further verify function.
Category 5
Well-known member
Hey Chunger/Matador. Looks good and I see you are indeed floating some high impedance parts for better long therm performance. Have you guys considered having a small switch with both the 47k and 100k R17 installed so tube types can be interchanged easily. The PSU is "close" to adaptable already with the 68K resistors. Dropping them slightly can make this more so, and your universal supply is already flexible enough thanks to regulation. Then you could use whatever tubes you have on hand, rather than having to get out the iron to make a switch. There are plenty of switches that are small enough to fit comfortably inside.
micaddict
Well-known member
Seems like a plan. 
Matador
Well-known member
Chunger hit most of the good points. Moisture absorption is better by a factor of ten, and impedance per cubic millimeter approaches that of PTFE. The same solder-mask formulation is also retained which performs nearly as good. This material is not something one would normally find in an audio project, so I wanted to try something new and expose folks to other options with state-of-the-art PCB formulations.
Nothing is really 'floated' that wasn't already before (with the exception of the tube grid pins, however this was more of an ease-of-assembly issue rather than one of isolation, as it was difficult to pack three wires into so tight a space and make everything easy to solder), but with this new material, there are no need for any teflon standoffs, so those are eliminated (and also reduce the BOM cost). Capsule connections are made right at the turrets now. HI-Z nodes are labeled with a circle in the pours so you can visually see which areas should be cleaned a second time. I was debating eliminating the soldermask completely, however the UV cured formulations are so much easier to clean with IPA that I decided to go ahead and keep it on there (the base board material being thermally hardened ceramic leaves a matte finish, and can be a real PITA to clean without ultrasonic methods).
Turret spacing is slightly wider as well, and there is no more need to run a wire from the plate cap back to the tube PCB: there is a new trace that does this. The prevents having to grind the turret to install the wire, and makes the whole thing easier to build and less prone to wire breakage.
Other than that, it all goes together the same.
Nothing is really 'floated' that wasn't already before (with the exception of the tube grid pins, however this was more of an ease-of-assembly issue rather than one of isolation, as it was difficult to pack three wires into so tight a space and make everything easy to solder), but with this new material, there are no need for any teflon standoffs, so those are eliminated (and also reduce the BOM cost). Capsule connections are made right at the turrets now. HI-Z nodes are labeled with a circle in the pours so you can visually see which areas should be cleaned a second time. I was debating eliminating the soldermask completely, however the UV cured formulations are so much easier to clean with IPA that I decided to go ahead and keep it on there (the base board material being thermally hardened ceramic leaves a matte finish, and can be a real PITA to clean without ultrasonic methods).
Turret spacing is slightly wider as well, and there is no more need to run a wire from the plate cap back to the tube PCB: there is a new trace that does this. The prevents having to grind the turret to install the wire, and makes the whole thing easier to build and less prone to wire breakage.
Other than that, it all goes together the same.
granger.frederic
Well-known member
- Joined
- May 22, 2010
- Messages
- 490
"Moisture absorption is better by a factor of ten, and impedance per cubic millimeter approaches that of PTFE"
Glad to see that you've changed your mind...You were strongly opposed to my remark the last time ...
but i still think that the cost benefit is too small to not use some Teflon standoffs, and the long term reliability will be more predictable (when a PCB is used)
Why not also rearrange all the components, to reduce to minimum all the distances, into the HiZ section ?
Glad to see that you've changed your mind...You were strongly opposed to my remark the last time ...
but i still think that the cost benefit is too small to not use some Teflon standoffs, and the long term reliability will be more predictable (when a PCB is used)
Why not also rearrange all the components, to reduce to minimum all the distances, into the HiZ section ?
Matador
Well-known member
granger.frederic said:
Ten times finer than fine is still fine. We're in the noise here for an encapsulated PCB assembly in both cases.
And since the entire PCB assembly matches the performance of PTFE (we're talking 1010 Mohms/cm here), adding more standoffs doesn't help.
granger.frederic
Well-known member
- Joined
- May 22, 2010
- Messages
- 490
Don't be as stubborn as a mule (for nothing) ...
You have understood the importance of long term moisture contamination, two post before...
Why not end the story with 2 or 3 teflon pins or floating point to point, like all serious professional mic manufacturers do...
the other slight improvement could be in reducing the HiZ distances.
You have understood the importance of long term moisture contamination, two post before...
Why not end the story with 2 or 3 teflon pins or floating point to point, like all serious professional mic manufacturers do...
the other slight improvement could be in reducing the HiZ distances.
chunger
Well-known member
granger.frederic said:
I suspect Matador rather enjoys introducing new, functional materials and processes to exceed the circuit's functional requirements and integrate these into his designs. Other people like to be "retro" and do very traditional layouts/materials.
This general design layout is flexible enough that you can teflon isolate as much of the circuit as you desire very easily.
granger.frederic
Well-known member
- Joined
- May 22, 2010
- Messages
- 490
Dear Chunger,
i need more pictures but if the teflon pins are only soldered through the holes (on the pcb) , it's useless.
the conductor of the pin must be "in the air" and the only the teflon must touch the PCB.
Teflon pins are only needed in the HIZ section, others can be metallic
still nothing about components rearrangement ...
In the past, Telefunken has improved the HiZ section in the ELAM by twisting the tube upside down...
You've already done that but why not make a complete redesign to reduce the distances even better ?
i need more pictures but if the teflon pins are only soldered through the holes (on the pcb) , it's useless.
the conductor of the pin must be "in the air" and the only the teflon must touch the PCB.
Teflon pins are only needed in the HIZ section, others can be metallic
still nothing about components rearrangement ...
In the past, Telefunken has improved the HiZ section in the ELAM by twisting the tube upside down...
You've already done that but why not make a complete redesign to reduce the distances even better ?
Matador
Well-known member
I'm more stubborn like a cat, not a mule, thank you very much.
Look: what you are describing I just don't think is a problem, and hasn't been for some time. This is DIY: I'm willing to instruct people to spend some extra time cleaning their PCB's, because the performance will be the same (perhaps even better), and there is no complicated build process. There are pages and pages of IPC's about heat-cured polymerized PCB coatings (like we use in this design) showing moisture absorption coefficients less than 0.05. The new material is 0.001. Both are fine. Optimizing a design for moisture absorption is desperately trying to solve a problem that just simply doesn't exist in this case. This isn't the 1960's any longer.
If I were making 100,000 of these, I might make different choices. In volume, my objective would be to minimize costs. If I thought floating a few components in the air might mean 0.01% less returns I might do it. I wouldn't pay someone in China $1 an hour to scrub PCB's with IPA because it would eat into my margin. But this is DIY: we don't have to do things by blind route.
You seem to be arguing (in this and other threads) that things just need to be a certain way, without defining the why. If you can prove that circuit interconnect impedances of 100+ petaohms in a tube mike is not sufficient, then lets discuss the circuit mechanism. Just because Neumann did it one way doesn't mean it's the best, just like I'm not arguing my way is the best. But we need to debating fact, not truth.
Look: what you are describing I just don't think is a problem, and hasn't been for some time. This is DIY: I'm willing to instruct people to spend some extra time cleaning their PCB's, because the performance will be the same (perhaps even better), and there is no complicated build process. There are pages and pages of IPC's about heat-cured polymerized PCB coatings (like we use in this design) showing moisture absorption coefficients less than 0.05. The new material is 0.001. Both are fine. Optimizing a design for moisture absorption is desperately trying to solve a problem that just simply doesn't exist in this case. This isn't the 1960's any longer.
If I were making 100,000 of these, I might make different choices. In volume, my objective would be to minimize costs. If I thought floating a few components in the air might mean 0.01% less returns I might do it. I wouldn't pay someone in China $1 an hour to scrub PCB's with IPA because it would eat into my margin. But this is DIY: we don't have to do things by blind route.
You seem to be arguing (in this and other threads) that things just need to be a certain way, without defining the why. If you can prove that circuit interconnect impedances of 100+ petaohms in a tube mike is not sufficient, then lets discuss the circuit mechanism. Just because Neumann did it one way doesn't mean it's the best, just like I'm not arguing my way is the best. But we need to debating fact, not truth.
granger.frederic
Well-known member
- Joined
- May 22, 2010
- Messages
- 490
Numbers,numbers....of what ?
give us the datasheet of your magic product or just the name...
i've never seen in my whole life a coated PCB that doesn't absorb a liquid with time, even if it's heat cured, light cured or everything you want cured .....
composite coating degrades (amongst other things) with time , moisture, air pollution and heat, period.
put your marvelous PCB material in a cup of tea during one month and look at it after....
if it's unaltered then run to your phone and call the NASA...
SO sad that are not in the 60's, that was the time when great people were building unrivaled and long lasting classic microphones....
Now in 2015 , other people are trying to clone them.
cheers
give us the datasheet of your magic product or just the name...
i've never seen in my whole life a coated PCB that doesn't absorb a liquid with time, even if it's heat cured, light cured or everything you want cured .....
composite coating degrades (amongst other things) with time , moisture, air pollution and heat, period.
put your marvelous PCB material in a cup of tea during one month and look at it after....
if it's unaltered then run to your phone and call the NASA...
SO sad that are not in the 60's, that was the time when great people were building unrivaled and long lasting classic microphones....
Now in 2015 , other people are trying to clone them.
cheers
Patrick from Davis
Well-known member
I kind of feel like Capt'n Obvious here, but this is the C12 Clone thread.
My Father had his own Custom PCB manufacturing shop for 30 years, and Teflon boards came to prominence in conjunction with Cell Phones. The microwave amplifiers required the dielectric to be very high for the cell tower equipment. Or that is what I remember from talking to him about it. The process to make those boards was very (quite annoyingly) difficult. You made plasma with nitrogen and hydrogen to "etch" the teflon, I guess tearing off the fluorines. Then it quickly had to be plated or the teflon polymers would turn over and become non reactive. Getting traces to have perpendicular edges was very painstaking as well.
When Matador mentioned that new epoxy versions that had similar properties to teflon, I can see why he jumped at using the material. If it plates like an epoxy board, then it is a drastically easier process, with consistent results, and a much reduced price. Apparently it can be produced so that it ends up even more sonically superior yellow as well. I am not sure why this discussion about the teflon has gone on this long. The turret version is available if you want to use teflon isolation. Otherwise, since Matador is designing this, and the material is a gain in performance, seems like a win to me. Or perhaps better put, a win for me. Also the only true way to compare is to build two in the various manners and swap the capsules and transformers between the two. I am guessing I won't be able to tell. At some point the gains in performance diminish to effectively nothing.
Patrick
My Father had his own Custom PCB manufacturing shop for 30 years, and Teflon boards came to prominence in conjunction with Cell Phones. The microwave amplifiers required the dielectric to be very high for the cell tower equipment. Or that is what I remember from talking to him about it. The process to make those boards was very (quite annoyingly) difficult. You made plasma with nitrogen and hydrogen to "etch" the teflon, I guess tearing off the fluorines. Then it quickly had to be plated or the teflon polymers would turn over and become non reactive. Getting traces to have perpendicular edges was very painstaking as well.
When Matador mentioned that new epoxy versions that had similar properties to teflon, I can see why he jumped at using the material. If it plates like an epoxy board, then it is a drastically easier process, with consistent results, and a much reduced price. Apparently it can be produced so that it ends up even more sonically superior yellow as well. I am not sure why this discussion about the teflon has gone on this long. The turret version is available if you want to use teflon isolation. Otherwise, since Matador is designing this, and the material is a gain in performance, seems like a win to me. Or perhaps better put, a win for me. Also the only true way to compare is to build two in the various manners and swap the capsules and transformers between the two. I am guessing I won't be able to tell. At some point the gains in performance diminish to effectively nothing.
Patrick
trans4funks1
Well-known member
- Joined
- Feb 4, 2013
- Messages
- 328
It seems like the newer PCB materials, that I have been reading about, are rated for dielectric properties while testing with 1gHz to 100gHz frequencies.
How do these newer products rate in, or near the audio spectrum?
How do these newer products rate in, or near the audio spectrum?
Matador
Well-known member
Patrick from Davis said:
Yes, exactly. Chunger did this exact experiment someplace earlier in the thread, if I can find it I'll post a direct link.
trans4funks1 said:
Yes, this was what they were originally developed for: mostly for cell phone/WiFi RF circuits that needed to transmit efficiently at 5GHz+, and also for networking backplanes in 10+GBit routing systems. Normal PCB material starts to look like dead shorts at these frequencies.
The previous solution was to ceramic laminate materials which were very expensive.For audio (more specifically high-impedance audio), the main benefit would be increased isolation properties between the different nodes. I've been wanting to try out this material for more dense designs (especially ones where there aren't turrets used), and the cost has finally come down to a point where it's just NRE for the fab to set up their line, so there we go.
Oh, and if you like to soak your microphones in tea, the performance is much better now. 8)
Category 5
Well-known member
Floated or not, moisture is bad for microphones, particularly capsules. Initiative One should be to store your valuable mics in a cool, DRY place. The bigger issue here is really just that there is a need to clean the board very thoroughly. It is flux, not moisture, that will possibly pose an issue. CLEAN THE BOARDS, and then clean them again. Isopropyl will not harm anything except for the polystyrenes, so clean, clean, clean.
As for moisture, keep one of these in each mic box. They can be re-activated in the oven at 250F and will make sure moisture will not be an issue for your mics.
http://www.amazon.com/Gram-Aluminum-Indicating-Silica-Canister/dp/B00BL79JDK/ref=sr_1_12?ie=UTF8&qid=1428074023&sr=8-12&keywords=silica+canister
As for moisture, keep one of these in each mic box. They can be re-activated in the oven at 250F and will make sure moisture will not be an issue for your mics.
http://www.amazon.com/Gram-Aluminum-Indicating-Silica-Canister/dp/B00BL79JDK/ref=sr_1_12?ie=UTF8&qid=1428074023&sr=8-12&keywords=silica+canister
Could someone please help me out; maybe Chunger, Matador or Dave.
I just received (2) C12 part kits from Chung {also with 251 options]; and have boards from Dave. But I don't seem to have several of the PS resistors that I see listed in the Official C12 BOM and, instead, have other odd value resistors; and instead of (4) 100uf caps there are (2) 100uf and (2) 47uf caps.
Is it possible that I'm looking at some early replaced version of the PS? I don't think I am because the parts in the BOM seem to match the parts in the schematics.
I don't have:
R1,R2 = 90.9k
R6 = 470k
R9 - 392 ohm
R10 = 1k
instead I have [and I do not think these are the microphone parts as they are in separate bags]:
Some of these values may be from meter readings and not necessarily the color codes.
(2) 400k
(2) 33k
(2) 68k
(2) 220 ohm
(1) 499k
Otherwise I have the (2) 1meg and the 4.7 ohm resistors.
I noticed that Dave's PCB uses 75k for R1,R2 where Matador's has 90.1k
And Dave's R4 uses a 250k trim after that, whereas Matador's uses a 100k.
I didn't look any further for any other differences because the other parts I first listed seem way off from what they should be.
Am I somehow making a mistake.
I just received (2) C12 part kits from Chung {also with 251 options]; and have boards from Dave. But I don't seem to have several of the PS resistors that I see listed in the Official C12 BOM and, instead, have other odd value resistors; and instead of (4) 100uf caps there are (2) 100uf and (2) 47uf caps.
Is it possible that I'm looking at some early replaced version of the PS? I don't think I am because the parts in the BOM seem to match the parts in the schematics.
I don't have:
R1,R2 = 90.9k
R6 = 470k
R9 - 392 ohm
R10 = 1k
instead I have [and I do not think these are the microphone parts as they are in separate bags]:
Some of these values may be from meter readings and not necessarily the color codes.
(2) 400k
(2) 33k
(2) 68k
(2) 220 ohm
(1) 499k
Otherwise I have the (2) 1meg and the 4.7 ohm resistors.
I noticed that Dave's PCB uses 75k for R1,R2 where Matador's has 90.1k
And Dave's R4 uses a 250k trim after that, whereas Matador's uses a 100k.
I didn't look any further for any other differences because the other parts I first listed seem way off from what they should be.
Am I somehow making a mistake.
